SA260324R-0112
① SA Source
- Source: 開啟完整 SA 文章
- Section:
GPU and LPU Integration: Attention FFN Disaggregation (AFD) - Line hint:
125
Context Before
Comparing the two operations, we see attention is stateful due to dynamic KV cache loading patterns, whereas
Evidence
FFN is stateless since the computation only depends on the token inputs.
Context After
Thus, we disaggregate the computation of attention and FFN. We map attention computations to GPUs, which handle dynamic workloads well. For FFNs, we map them to LPUs, since LPU architecture is inherently deterministic and benefits from static compute workloads.
② Atomic Claim
FFN 是 stateless，因為其運算只依賴 token input。
- Epistemic Mode:
ASSERTED - Mapping Status:
COMPLETE
③ Semantic Frame
{
"attribute": "STATEFULNESS",
"context_nodes": [],
"entity": {
"id": "04_knowledge_base/Feed-forward Network",
"label": "FFN"
},
"frame_type": "ATTRIBUTE",
"qualifiers": {
"condition_text": null,
"numeric_mentions": [],
"temporal_mentions": []
},
"value": {
"numeric_mentions": [],
"value_text": "stateless；運算只依賴 token input。"
}
}④ Canonical Entity Mapping
| Role | Surface Label | Canonical Target |
|---|---|---|
| entity | FFN | 04_knowledge_base/Feed-forward Network |
⑤ Human Review
請在 Properties 逐項確認：
- 原文 → Atomic Claim 是否忠實
- Atomic Claim → Semantic Frame 是否忠實
- Canonical Entity mapping 是否正確
- Epistemic mode 是否保留原文語氣
- 最後選擇
review_action
Review state
Markdown 內文不是正式 approval。只有 Apply bridge 寫入的 Decision Ledger event 才是正式決策。